A service backlog is the volume of pending work not yet completed. The most direct operational KPI among the options that influences backlog is service lead time (how long it takes to complete a service request from initiation to closure). When lead time increases, work stays open longer, and the backlog tends to grow unless intake demand falls. Conversely, reducing lead time (through better scheduling, capacity planning, process improvements, or fewer reworks) helps clear work faster and prevents backlog accumulation. Revenue is typically an outcome financial KPI and does not directly control queue volume. “Service complaints responded to (%)” reflects responsiveness to complaints; it may improve satisfaction but does not necessarily reduce total pending service volume. “Productivity” is too vague as written (it needs a clear formula, e.g., jobs completed per technician per day) and therefore is less defensible than a specific cycle-time measure. In KPI systems, backlog is usually balanced with capacity and flow KPIs (arrival rate, completion rate, lead time). Measurement challenges include consistent definitions of “open” vs “closed,” and ensuring timestamps are accurate to avoid distorted lead-time and backlog analysis.